Welcome to the Best SEO Blog!


The latest in search engine marketing tactics, the tried and true techniques. Feel free to comment or suggest topics that you would like to know more about.

January 26 2009

Beware the SEO mentor trap

If you’re going to contribute content to other Web sites, the best thing you can do is NOT offer them advice on how to improve their content, on how to market themselves, and how to optimize for search.

People like Larry Page and Sergey Brin have proven beyond all reasonable doubt that anyone can become a billionaire regardless of how little they actually understand the World Wide Web (and Page and Brin didn’t have a clue when they proposed PageRank, which ranks among the Ten Most Stupid Ideas On The Internet).

In other words, you may know better than the other guy how to “market” his stuff, but his passion, ambition, connections, and just plain dumb luck may accomplish more than you can ever hope to do for him. If Larry Page and Sergey Brin had asked for and followed advice from the SEO community they would be out of business today. Everything they’ve done with Google has flown in the face of conventional search engine optimization community wisdom.

Their front page is ugly, but ugly works.

They refrained from writing actual content for several years, and instead replicated everyone else’s content.

They waited a long time to monetize their site.

They left their site and content open to clear and obvious manipulation for years (in fact, it’s still easy to manipulate Google’s search results).

If Larry Page and Sergey Brin were to be judged for their search engine optimization today, they would be considered miserable failures and excruciatingly naive spammers. In fact, Google recently achieved the distinction of ranking third on Spamhaus’s list of top spam resources.

My point here is that if you’re going to participate in someone else’s Web site by contributing content, then the best thing you can do for yourself is to keep your advice to yourself. Unless the site operator pays you money for advice, don’t share it. Why? Because he may just prove to have the next best thing and he won’t need your advice.

100% of all SEOs offer ineffective advice at least some of the time. No one escapes from the curse of Being Inconveniently Wrong in this industry. That’s just the way it is.

Free advice is usually the worst advice. It’s given on the fly, with little to no research or consideration, and no real actual knowledge of what the Web site’s goals are. Maybe the site operator doesn’t know what his goal is. All he knows is he has a hot idea that he wants to work.

You should decide whether to contribute content on the basis of whether you like his site and not for any other reason. You should not be trying to boost his brain power. He may be smart enough (most people say Page and Brin are pretty smart guys). He may also be challenging some marketing assumptions. And his approach may work better than your tried and true methods.

If you contribute to someone else’s site, just let him run with your content and see what happens. You may find he can help you more than you can help him.

Written by Michael Martinez

January 14 2009

Building a Bot Accessible Archive

Tagged Under : , , , , , ,

Bot accessibility is about making sure that search agents known as spiders, robots or “bots” for short, have full accessibility to all the content that you want displayed to the public.

Creating a growing archive that remains accessible to both bots and the public can be instrumental to rapidly increasing the size, strength and overall value of your web property. Like most things in SEO, it is much easier to plan ahead for an effective archive rather than reorganizing one that has been indexed.

Rather than focusing on which database to use or how to build with one of the many excellent content management systems out there, I’m going to list a few key points to watch for.

Bot Accessible Archive Dos:

Each Document Should Have Its Own URL
Whether using pages, posts (or anything else) make sure that each article/press release/product has its own URL. Giving each subject its own URL is probably the most important aspect of proper archiving.

Persnickety Order
Create a logical hierarchy and categorize groupings of similar subject matter. Clean informative structure is as much about human accessibility as it is about bot accessibility.

Proper Labeling and Legible URLs
On the server side, convert all database generated URLs to descriptive URLs. The actual logistics of this have everything to do with your CMS and platform. Labeling the tags and content in your archive in a way that is human intuitive will allow indexing bots to serve your pages to the correct searchers.

Using legible URLs can convey on the SERP level what exactly is on a page and may even lend to your target keywords. Searchers will have to rely less on your descriptions and title tags to surmise what your page is about.

Example:

Bad Archive Structure

Good Archive Structure

Bot Accessible Archive Don’ts:

Flash Archives
Yes I have seen Flash archives that were pretty to look at but overall somewhat cumbersome to navigate and not necessarily bot accessible. Since my guess is that you aren’t trying to impress anyone with your creative prowess as they rummage through your archive, it’s best to apply KISS and focus on functionality.

Ajax Archives and iFrame Archives
Both Ajax and iFrames tend to have a single URL and display content through a window on the page. The problem is when you navigate to the press release or document that you are searching for there is virtually no way to share it without downloading or copy and pasting it.

PDF Archives
While it is possible for a bot to crawl and index a PDF document, people generally hate to see PDFs in search results unless they are searching specifically for a product manual or academic paper. PDFs make good product brochures and lousy product pages.

PDFs appear to pass value such as anchor text, but from a user intuitive standpoint can make awkward link placement. A user clicks on a PDF, it downloads, a link is visible so they click it and the browser opens a new window. Ick!

If your archive happens to be of product manuals or academic papers you might want to consider creating a database driven archive of pages containing an abstract, author and a way to download the PDF.

Happy archiving!

Written by Nicholas Ramirez

January 12 2009

Keep your backup plan current

For several years I have used an email address for Eudoramail.com (now mail.lycos.com) as a backup to my own personal email addresses. I’ve changed ISP providers every time I have moved between states so I’ve never had a stable email address outside of those I provided myself through Xenite.Org (which I’ve had to change twice due to spam).

In other words, at any given time I usually have two active email addresses: my Xenite address that I use for almost everything and my Eudoramail address that I’ve used for only a very few things, as well as a backup in case Xenite had problems.

For several months last year, Xenite had email problems and I found myself relying on the Lycos account more and more.

This weekend, however, I could not reach Lycos.com. None of their services seemed to be online. I searched for news and blog articles about problems at Lycos and found nothing except stories from last month about Lycos Europe going out of business. Lycos Europe is supposed to be different from Lycos in the United States, but Lycos Europe had told all its Web hosting clients to find new providers before their sites went offline early this year.

As I write this Lycos.com is back online but the outage came at a most inconvenient time, as several domains which I had registered using the Eudoramail account were about to expire. In the process of moving those domains to a new registrar I found I could not approve the transfer requests because I could not log in to Lycos.

The quick resolution was to change the administrative email so that I could finish the transfers. But given the current state of the economy, I have been wondering if we’re not about to see another dot-com meltdown. I’ve already seen many AdSense publishers complaining about a decline in revenues. I have to admit that the decline has affected me, too. The Xenite network has other sources of revenue so we’ve still got money in the bank but the downturn in revenues has come at an inconvenient time.

Which leads me to the object lesson for today: not always do you want to have a backup plan, you want to make sure your backup plan is current and relevant. I’ve had to rely on two backup plans this month: one for email and one for revenue. If I actually lose either the Lycos account or if AdSense just stops being a useful source of income, I’m going to have to develop one or two more backup plans.

As for email, there are still some alternatives out there but none of them seem palatable. Gmail, for example, is rife with system problems if all the complaints people post are to be believed — and people like me have begun blocking Gmail because of all its spam issues (Spamhaus currently lists Gmail as the third most active source of spam).

Remonetizing a network is not so easy. Advertisers are cutting back on expenses just as consumers are cutting back on purchases. I’m exploring some options before it’s too late. If you’re monetizing Web sites and you’re not sure of what’s going on, trust me: there is a HUGE contraction in process. I’ve seen discussions in many forums and concerned posts on many blogs since the beginning of December.

The majority of gung ho “everything is okay” commentary I’ve seen has come from people pushing yet another “get rich quick” affiliate program on their blogs. I got through the last dot-com meltdown by not relying on Web revenues. I expanded my content network and experimented with new monetization schemes. But truth be told, I went through a 2-year spell where everything I tried failed to pay off. I never thought AdSense would last forever but I have been hoping something else would come along before AdSense began sputtering out.

Written by Michael Martinez

January 08 2009

Why you need descriptive text for internal links

Rand Fishkin just published the results of an internal link anchor text test on SEOmoz. He doesn’t provide much information so we cannot judge the validity of the test (although his selected keyword, “SEO”, is highly competitive).

As I have said many times, SEOs put way too much emphasis on links and link anchor text. The SEO community follows a self-fulfilling prophecy about links: “I believe I need links to rank, therefore I will only try to rank on the basis of links, therefore links matter more for rankings than anything else.”

So let’s take our SEO tin foil hats off for a moment and look at link anchor text the way we should when we’re not trying to improve our search rankings. When you arrive on a Web site through an entry page and you don’t know what the site is about, do you want a clue or not?

Link anchor text can provide very helpful visual cues. So can the text preceding link anchor text.

Ian McAnerin suggested a way for people to organize links that incorporates context-setting text. Instead of simply providing three links for “Gray Widgets”, “Blue Widgets”, and “Brown Widgets” in a hypothetical navbar, he suggests the navbar could look more like this:

I am looking for a: Gray Widget | Blue Widget | Brown Widget

Regardless of whether the link anchor text is going to help the destination pages in search results, it’s definitely going to help the visitor. And I submit to you that whatever you call your pages internally, as long as it is meaningful and relevant to the page content, will influence how your visitors link to those pages, too.

Some people will use inexpressive descriptors in their anchor text to avoid passing value, but most people will refer to your page the same way you do (or very closely to the way you do) if you use concise, useful, descriptive anchors or meaningful page titles.

The link anchor text can also help make the linking page more relevant to its own topic if it is linking out to similar pages. I’ve often used lists of links to make pages more relevant in elite directory pages. Just because you embed keywords in link anchor text doesn’t mean those keywords won’t help the page on which they are found.

Search engines may or may not care about how your internal link anchor text appears. I can think of many situations where it doesn’t matter to a search engine. But if you place a link on a page it needs to be useful, helpful, and informative.

Written by Michael Martinez

January 05 2009

Things It Is Good To Know

Not every aspect of search engine optimization is about your Web sites, your visibility, your traffic, and your conversions. There are some very naughty, unscrupulous people out there who will walk all over you if you allow them to. They do this in their ongoing efforts to cheat the search engine community.

Yes, friends, I’m talking about those so-called Black Hat Search Engine Spammers.

There are some things members of the black hat community do that I just don’t appreciate, agree with, or tolerate (when they get in my face). Nor should you tolerate their nefarious activities when these anti-social behaviors abuse your resources. Here is a list of things you should know.

Script Kiddies Rape Your Registrations – If you operate a forum, a blog, or some other UGC Web site where people can register and post content, you’re all but hanging a “Spam me” sign on your back. The script kiddie link spammers (most of whom in my experience promote adult content, illegal pharamaceutical content, and transparent affiliate sites) will hammer your server with bogus registrations. Once registered, they will hammer your server with comments, posts, and other link-filled user-generated content.

You can counteract these jerks’ best efforts by:

  1. Requiring confirmation of all registrations (force them to respond to confirmation emails)
  2. Moderating all registrations and first time posts (some scripts will post up to 4 or 5 random linkless comments before they start dropping links)
  3. Blocking all Russian, Ukrainian, Chinese, and other non-English domains
  4. Blocking hyperactive IP addresses in your “hosts.deny” and/or .htaccess file (or equivalent)

There ARE people in Russia and the Ukraine (and other eastern European nations) who may want to visit your forums, but if you see 100 registrations in one day from accounts like “asgheod@menshealthusa.ua” it’s a pretty safe bet a script kiddie is trying to abuse your service.

Some people favor using CAPTCHAs over confirmation emails. I actually use both. Scripts have been written that actually get past typical CAPTCHAs. How do they do it? I don’t know. They just do.

Of course, Google claims some people in India sell their registration services. These people sign up for Gmail accounts and other free email accounts. I’m currently blocking Gmail and Hotmail from my own forums because of abuse from their users. So the listen to take away here is: When you use free email services, you look like a spammer.

Your images appear in the strangest places – If you post interesting pictures or pictures of celebrities and politicians on your Web site, the odds of those images appearing in other peoples’ forums and blogs are astronomically high. Over the past couple of years many people in the SEO community have noticed increased search referrals from image search, and there has been growing interesting in optimizing for image search.

But here’s the trick: image search referrals rarely convert well for most sites. Those referrals are usually people looking for images to share in forums and blogs. They tend to hotlink to your images and eat up your bandwidth.

Here are a few ways to counteract hotlinking:

  1. Restrict access to your images to “known” domains. That includes your domains and search engine domains.
  2. Substitute a standard promotional ad for all hotlinked images.
  3. Brand all your images with your domain name.
  4. Block image search from indexing your images.

If you cannot think of a good reason to be found in image search, then just don’t fall into the image search trap. Many people do block the search engines from indexing their images. But there is no hard and fast rule on this. Consumers often use image search to find specific products (particularly where product names are ambiguous, like “Rolex watch” or “Citizen watch”).

If you do want to leverage image search to draw traffic to your site, then make sure that you discourage hotlinking. You cannot prevent people from capturing images and reusing them (even if you go to extreme lengths, all they have to do is take a snapshot of their screen and they have your image). Image search referral can bring customers to your site, but it will more likely bring bandwidth thieves. Make it hard for them to abuse your bandwidth.

Not all email packages are alike – Believe it or not, the old UNIX sendmail service can be more easily secured than some of today’s popular “safer” email packages. I’ve written about Qmail spam exploits in the past. Qmail’s creator claimed it was not vulnerable to spam exploits. The spammers proved him wrong.

Postfix is another email package that is supposedly better than sendmail. However, the braindead designers of Postfix decided it was better to not allow you to blacklist and whitelist domains with “hosts.allow” and “hosts.deny” (which you can do with both sendmail and Qmail). If you go through enough Rube Goldberg machinations, you can sort of filter spam with Postfix but the learning curve and implementation time are ridiculously high compared to older, “less secure” sendmail and other packages.

Internet Service Providers combat email spam in several ways that may affect you regardless of which email package you use.

  1. They detect and block open relays.
  2. They perform reverse DNS lookups and reject emails from servers that fail these tests.
  3. They subscribe to the wrong black lists.
  4. They white list domains and require you to ask to be included in the white list.

Not all emails are bounced back when they fail these tests. Theoretically, the ISPs are SUPPOSED to send a bounce message, but some ISPs refuse to do that. These rogue admins either believe you’re a spammer because someone says you are a spammer and therefore you should be ignored like a spammer OR they just don’t have a clue about how to handle spam email.

There are some reputable blacklist services (like Spamhaus.org) that allow you to find out your server is an open relay, fix the problem, and then be removed from their black lists. Then there are guys like Joe Jared who don’t believe in treating people fairly. Jared’s blacklists have been known to block as many as 32,000 IP addresses at a time with no option for vetting or whitelisting individual domains and IP addresses. He and his friends historically pursued economic blackmail against hosting providers whom they concluded were “spam-friendly”. By driving customers away from those services, Jared and his friends thought they were doing themselves and everyone else a favor.

A spammer took down Jared’s blacklist at his OsirusSoft site a few years ago. I don’t know if he is currently engaged in blackmailing hosting providers with another blacklist, but the day Jared’s site went day was the one day in history I was applauding an email spammer. I don’t encourage or condone email spam — but Jared’s solution was worse than the cure and it hurt a lot of people, forcing us to move our domains more than once (an expense Jared and his cronies did not care to share or ameliorate in any way).

If your server is failing reverse DNS lookups despite your best efforts to fix the problem, check with your hosting provider. Many Web host services now routinely control email DNS settings at their level to help fight email spam. They will usually work with clients who need to pass reverse DNS lookup. Don’t be rude and angry with your host for failing to disclose this security practice to you. Just be glad they are trying to stay out of the gunsights of idiots and morons like Joe Jared and his friends.

I have found a growing number of ISPs (like AT&T and all their subsidiaries) require human review for domains or IP addresses that fail spam tests. Although it is inconvenient to have to fill out their forms and wait for the review process, I am grateful to them for taking the time to work with Webmasters on these issues. They’ll remove blocks when you show that you’ve closed your open relay or have resolved your reverse DNS lookup issue.

But be warned: if your site is repeatedly hacked or exploited you may find it elevated to a tighter restricted list, and getting off that won’t be so easy. We may be coming to the day when it’s better to allow a reputable third party (or your hosting ISP) handle your domain’s email.

But I’m not ready to hand the reins over to Gmail just yet. I’ve heard some Web sites actually block their emails because of spam abuses.

Take that for what it’s worth.

Written by Michael Martinez